Comparison of Results of Mussel Watch Programs of the United States and France with Worldwide Mussel Watch Studies

Transcription

1 Pergamon PII: S25326X(98)49-6 Marine Pollution Bulletin, Vol 36, No 9, pp , Elsevier Science Ltd All rights reserved Printed in Great Britain X/98 $19 + Comparison of Results of Mussel Watch Programs of the United States and France with Worldwide Mussel Watch Studies A Y CANTILLO Office of Ocean Resources Conservation and Assessment, NOAA/National Ocean Service, 135 East West Hwy, Silver Spring, MD 291, USA As part of the Global Ocean Observing System (GOOS), the National Oceanic and Atmospheric Administration (NOAA) National Status and Trends () Program compiled the World Mussel Watch (WMW) data base with results from the analyses of marine or estuarine mussels or oysters as far back in time as possible Here we compare WMW data with results from two long-term Mussel Watch Programs, the R~seau National d'observation de la Qualit~ du Mulieu Marin () Mussel Watch in France and Program in the United States The medians and 85th percentiles for Cr, Ni, Cu, Zn, As, Se, Ag, Cd, Hg and Pb in mussels and oysters were calculated for the WMW, and Mussel Watch programs While there was generally good agreement for medians among all three data sets, the upper ends of the WMW concentrations tend to be higher than their and counterparts This probably reflects the fact that the latter two programs emphasize collection of mollusks at representative sites rather than within small areas of extreme contamination such as near waste discharges 1998 Elsevier Science Ltdo All rights reserved In response to increased awareness off possible global environmental changes, the Global Ocean Observing System (GOOS) was created under the auspices of the United Nations (Anderson, 1997) as a framework to put national data into a global context Mussel Watch programs have been used to assess coastal environmental pollution since they were proposed in the 197s by Goldberg et al (1978) In the spirit of GOOS, the National Oceanic and Atmospheric Administration (NOAA) National Status and Trends () Program undertook the task of providing the global context for these types of data by compiling results of Mussel Watch programs throughout the world World Mussel Watch Data A worldwide literature search for reports with results from chemical analyses of soft parts of marine or estuarine mussels or oysters yielded more than 3 o El "o "6 $ Q E ~3 Z 7 6, t T -'' MUSSELS OYSTERS ~ Year Fig 1 Histogram of years for which data are available for mussels and oysters in the WMW database 712

2 Volume 36/Number 9/September 1998 W I 18 I I 12 6 i I if,",--' ",, l =i = 3" " ~=% i= ==l -- E I I I r= It u" -6-9 S Fig 2 Geographical distribution of all sampling locations in the WMW database citations Most of the reports were based on quantifying contamination, but some of the oldest were on the nutritional use of oysters in the treatment of anemia Results for the trace elements Cr, Ni, Cu, Zn, As, Se, Ag, Cd, Sn, Sb, Hg, Ti, and Pb in mussels and oysters were entered into four files to create the World Mussel Watch (WMW) data base The trace organic contamination data were much fewer than for trace elements and are difficult to compare among sources Often, insufficient analytical methodology detail was available in citations to judge to what degree one set of data could be compared to another There have also been changes in analytical methodology for the analysis of trace organic contaminants in the environment over the last two decades so comparison of older data with more recent data even from the same source may not be valid It was decided, TABLE 1 Countries or territories represented in the WMW database Argentina Hong Kong Peru Australia Iceland Poland Bahrain India Portugal Belgium Indonesia Puerto Rico Bermuda Ireland Russia Brazil Isle of Man Scotland Canada Italy Singapore Chile Japan South Africa Columbia Korea Spain Croatia Kuwait Sweden Denmark Lebanon Taiwan Egypt Mexico Thailand England Monaco The Netherlands Finland Morocco United States France New Zealand Wales Germany Northern Ireland Yugoslavia Greece Norway Greenland Oman Q_ m v "5 4-[ E Z 35, Arithmetic scale WMW median = 13 gg/g dry wt 4 5 MUSSELS Zn WMW 85th percentile = 26 gg/g dry wt v' o o Logarithmic scale Concentration (pg/g dry wt) Fig 3 Distribution of zinc in mussels on arithmetic and logarithmic scales (WMW mean is the mean of the logarithmic values and the WMW high is the mean plus one standard deviation of the logarithmic values All data used in calculations Some high concentrations not shown) 713

3 Marine Pollution Bulletin therefore, not to include the trace organic contaminant levels in the WMW Results of the analysis of trace metals in more than 6 species of mussels and oysters, mostly Mytiloids and Ostreoids, were entered into four files to create the WMW database: TM, DRY, TM WET, TM RANGES DRY and TM RANGES WET Data from citations containing trace metal (TM) concentrations on a dry weight basis were entered into the TM DRY file If there was sufficient information in the citation to convert wet weight results to dry weight, the concentrations were converted and added to the TM DRY file The TM WET file contains only those concentrations reported on a wet weight basis without information on wet-dry ratios In some citations, only the mean, low, and high concentrations were reported and these are contained in the TM RANGES DRY or TM RANGES WET files In addition to the trace metal data, each file contains information on data source, latitude and longitude of sampling sites, sampling location name, and country OYSTERS [ Zn Values which were listed as below a specific detection limit were listed at that level (eg <5 was listed as 5) In cases where mollusks were transplanted, only concentrations measured prior to transplanting were used No attempt was made to obtain approximate concentrations from graphical data displays although this may be possible in many cases If no sampling date was listed in the data source, the publication year of the citation was used If a period of time was listed, then the time half-way between the two stated dates was used Cantillo (1997) includes a summary of each data source in terms of: bibliographical citation; country; year of sampling or publication; number of sites sampled; species sampled; a short description of the analytical methodology; a notation of the use of reference materials or participation in intercomparison exercises; collection and analysis of other species or other types of samples such as sediment or seawater; and country, author, species and analyte indices The data of the NOAA and the R6seau National d'observation de la Qualit6 du Milieu Marin () Mussel Watch efforts were not added to the 5 4 MUSSELS and OYSTERS Cd c O D "O *6 Q z,,,j~ Arithmetic scale WMW mean = m OYSTERS 16 gg/g ~ ~ Zn dry wt WMW 85th percentile = 45 p_g/g dry wt t'~ 1 "o "6 5 J~ E Z Logarithmic scale Concentration (gg/g dry wt) Fig 4 Distribution of zinc in oysters on arithmetic and logarithmic scales (WMW mean is the mean of the logarithmic values and the WMW high is the mean plus one standard deviation of the logarithmic values All data used in calculations Some high concentrations note shown) Logarithmic scale Concentration (ltg/g dry wt) Fig 5 Distribution of cadmium and copper in musselsand oysters on arithmetic and logarithmic scales (Oysters and mussels differ in their ability to concentrate some elements such as Cu) 714

4 Volume 36/Number 9/September 1998 WMW database The results of these programs were treated separately since the numnber of data points in each is larger than that of the WMW data The Mussel Watch data can be obtained from the Program Office by contacting the Manager, Program, NOAA/NOS/ORCA 21, 135 East West Hwy, Silver Spring, MD 291, USA, or from the Internet at: the data can be obtained by contacting D Claisse, IFREMER, BP 115, Nantes Cedex 3, France WMW Database Statistics Time and geographical distribution Data found for samples collected before 194 (Fig 1) were mostly from studies of the use of oysters in the treatment of anemia No data were found for the decade of the 194s and few prior to 197 The large increase in data in the 197s reflects the beginning of the environmental movement and the inception of 'Mussel Watch' monitoring by Go![dberg et al (1978) The distribution of data by latitude and longitude shows that data were collected throughout the world (Fig 2) However, in the listing of countries (Table 1), the sparcity of data from the continents of South America and Africa is noted Statistics per element Oysters and mussels are not equal in their ability to concentrate trace elements (O'Connor, 1993) The trace elements Ag, Cu, and Zn are enriched more than 1-fold in the oyster C virginica relative to the mussel M edulis Data for other metals do not show so strong a species effect Nonetheless, there were many species of mussels and oysters used in the compilation, so it was decided to simplify the initial evaluation of the WMW data base by treating the mussel and oyster data separately The most commonly determined elements in mussels and oysters were Cu, Zn, Cd, Pb, Hg, Cr, and Ni Examples of the Zn data distribution with time and histograms are shown in Figs 3 and 4 The Cr, Ni, Cu, Zn, Cd, and Pb data for mussels and oysters were found to be log normal The As data also appear to be log normal although the distribution may be bi-modal There are some Cu and Zn data dating back to the turn of the century The levels found back then were not the highest found in the literature, although the specimens analyzed were usually 'green' oysters with known high levels of Cu No data were found for Se in oysters and the very little data found for Se in mussels appear to be log normal within a small concentration range The Ag and Hg data distributions do not clearly TABLE 2 Comparison of WMW,, and median and 85 percentile concentrations (gg g- 1 dry wt) a'b Mussels Oysters Median 85th percentile n c Median 85th percentile n c Cr WMW Ni WMW Cu WMW Zn WMW As WMW Se WMW Ag WMW Cd WMW Hg WMW Pb WMW ~The 85th percentile concentration is the one below which are the concentrations of 85% of the samples The median concentration is the 5th percentileball data used in calculations data cover 1986 through 1995 data cover 1979 through 19945n = Non-zero values in the data sets 715

5 Marine Pollution Bulletin TABLE 3 Comparison of WMW,, and median and 85 percentile combined mussel and oyster concentrations (lag g- J dry wt) a Median 85th percentile n Cr WMW Ni WMW As WMW Cd WMW Hg WMW Pb WMW ~Footnotes as in Table 2 do not require the assumption that the data be log normal Table 2 lists the medians and 85th percentile concentrations ('highs') for trace metals from the WMW,, and data sets There was generally good agreement for medians among all three data sets The medians and 'highs' appear to be the lower of the three sets of data Beliaeff et al (1997) compared the and data in detail and explained differences in national medians in terms of different patterns of metal use The WMW 'highs' tend to be higher than their and counterparts This is probably due to the fact that the latter two programs emphasize collection of mollusks at representative sites rather than within small areas of extreme contamination such as near waste discharges Within the WMW compilation, on the other hand, some data came from programs specifically designed to sample 'hot spots' and these are evident at the upper end of the concentration distributions show log normal distributions Low levels and analytical methodology could contribute to this data distribution pattern Most of the data found for Sn resulted from measurements of tributyltin rather than elemental Sn and few data were found for Sb and TI No conclusions could be drawn for these three elements Comparison with long-term mussel watch programs O'Connor and Beliaeff (1995) summarized data using geometric means and geometric means plus one standard deviation to define 'high' concentrations This was based on the log normal nature of the data distributions and the fact that the concentration range beyond one standard deviation from the mean contained about 17% of the data A simpler way to summarize the data is to use medians and 85th percentiles These in turn are similar for the data and TABLE 4 Concentrations of trace metals in mussels and oysters that are indicative of contamination (lag g i dry wt) Metal Concentration Mussels and oysters Cr 25 Ni 34 As 16 Cd 37 Hg 23 Pb 32 Mussels Cu 1 Zn 2 Ag 75 Oysters Cu 3 Zn 4 As 5 Compilation of all data for mussels and oysters As previously discussed, mussels and oysters are not equal in their ability to concentrate trace elements For example, the WMW concentration histograms for Cd and Cu show a bi-modal distribution for Cu due to species difference between oysters and mussels for Cu uptake but no such difference for Cd (Fig 4) The differences for Cr, Ni, As, Se, Cd, Hg and Pb were not sufficiently high as to prevent the combination of the mussel and oyster data and the mussels and oysters in each of the WMW, and databases were combined and the medians and 85th percentiles calculated (Table 3) No Cr, Ni or As data were available from There were insufficient Se data in the WMW database to use meaningfully The agreements among the medians of the three programs for some elements are very close Some differences such as the high median and 85th percentile for Pb and high 85th percentile for Cd in the WMW remain Lower Pb levels in the data can be explained by the history of use of these metals (Beliaeff et al, 1998) The main conclusions remain in that differences among medians for the three data sets are relatively small while the WMW data set has higher 'highs' Conclusion There were few data from common locations and several years so no determinations of worldwide temporal trends have been made Few data were found prior to the 197s During the 197s, there was a significant increase in the number of citations reporting the use of mussels and oysters to determine contamination level and monitor coastal areas Many of these studies included highly contaminated sites, resulting in very high trace metal values The high values reported 716

6 Volume 36/Number 9/September 1998 during the 197s may be the result of increased awareness of coastal contamination and evaluation of contaminates sites, not necessarily of increased pollution levels In the absence of human actmty, there would be measurable concentrations of trace: elements in mussels and oysters We have no independent basis for declaring any specific concentration as a natural limit, but to a first approximation, we could use the lowest of the 85th percentiles from the or Mussel Watch programs as indicative of contamination (ie elevated by human activity) The WMW 85th percentiles include data from heavily contaminated sites and thus are not comparable to and percentiles These concentrations (Table 4) in any mussel or oyster could denote contamination There are mollusks with lower concentrations that are contaminated Conversely, there are sites where local mineralogy causes mollusks to have greater concentrations of one or another of these trace elements and are not contaminated Nonetheless, as a first approximation, these limits can be used by GOOS for classifying concentrations reported by monitoring programs from around the world Even without classifying, new data can be put into their worldwide context by comparing them with the data summarized here or by inserting them into the WMW database The author wishes to thank the chemists that produced the data, over many years, which were used to compile the WMW database, the Staff of the NOAA Central Library for their assistance in locating numerous citations T O'Connor for his suggestions, and D Harris for scanning, proofreading, and his many, many trips to the library Anderson, N (1997) An early warning system for the health of the oceans Oceanography 1, Beliaeff, B, O'Connor, T P and Claisse, D (1998) Comparison of chemical concentrations' in mussels and oysters from the United States and France Environ Monit Assess 49, Cantillo, A Y (1997) World Mussel Watch Database NOAA Tech memo NOS ORCA 19 NOAA/NOS/ORCA, Silver Spring, MD 179 pp Goldberg, E D, Bowen, V T, Farrington, J W, Harvey, G, Martin, J H, Parker, P L, Riseborough, R W, Robertson, W, Schneider, E and Gamble, E (1978) The Mussel Watch Environ Conserv 5, O'Connor, T P (1993) The NOAA National Status and Trends Mussel Watch Program: national monitoring of chemical contamination in the coastal United States In Environmental Statistics, Assessment and Forecasting, ed C R Cothern and N P Ross Lewis PUN, Boca Raton, FL O'Connor, T P and Beliaeff, B (1995) Recent trends in coastal environmental quality: results from the Mussel Watch Project NOS ORCA Spec Pub NOAA/NOS/ORCA, Silver Spring, MD 4 pp 717